Download Keystone species - Department of Conservation

Keystone species: the concept
and its relevance for conservation
management in New Zealand
SCIENCE FOR CONSERVATION 203
Ian J. Payton, Michael Fenner, William G. Lee
Published by
Department of Conservation
P.O. Box 10-420
Wellington, New Zealand
Science for Conservation is a scientific monograph series presenting research funded by New Zealand
Department of Conservation (DOC). Manuscripts are internally and externally peer-reviewed; resulting
publications are considered part of the formal international scientific literature.
Titles are listed in the DOC Science Publishing catalogue on the departmental website http://
www.doc.govt.nz and printed copies can be purchased from [email protected]
© Copyright July 2002, New Zealand Department of Conservation
ISSN 1173–2946
ISBN 0–478–22284–X
This report was prepared for publication by DOC Science Publishing, Science & Research Unit; editing
by Lynette Clelland and layout by Ruth Munro. Publication was approved by the Manager, Science &
Research Unit, Science Technology and Information Services, Department of Conservation, Wellington.
CONTENTS
Abstract
5
1.
Introduction
6
2.
Keystone concepts
6
3.
Types of keystone species
8
3.1
3.2
3.3
3.4
4.
Organisms controlling potential dominants
Resource providers
Mutualists
Ecosystem engineers
8
10
11
12
The New Zealand context
14
4.1
4.2
4.3
4.4
14
16
18
19
Organisms controlling potential dominants
Resource providers
Mutualists
Ecosystem engineers
5.
Identifying keystone species
20
6.
Implications for conservation management
21
7.
Acknowledgements
22
8.
References
23
4
Payton et al.—Keystone species: the concept and its relevance in New Zealand
Keystone species: the concept
and its relevance for conservation
management in New Zealand
Ian J. Payton 1, Michael Fenner 2, William G. Lee 3
1
2
3
Landcare Research, P.O. Box 69, Lincoln 8152, New Zealand
Biodiversity & Ecology Division, School of Biological Sciences, University of
Southampton, Southampton, SO16 7PX, United Kingdom
Landcare Research, Private Bag 1930, Dunedin, New Zealand
ABSTRACT
The keystone species concept has proved both promising and elusive in theoretical
and applied ecology. The term has its origins in Robert Paine’s studies of rocky
shore communities in California. When the top predator (a starfish) was removed
the species assemblage collapsed, prompting the architectural analogy with the
keystone of an arch. By definition keystone species are those whose effect is large,
and disproportionately large relative to their abundance. They include organisms
that (i) control potential dominants, (ii) provide critical resources, (iii) act as
mutualists, and (iv) modify the environment. Identifying keystone species can be
problematic. Approaches used include experimental manipulations, comparative
studies, natural history observations, and ‘natural experiments’, but no robust
methodologies have been developed. Our inability to monitor and manage all
aspects of biodiversity has led to the development of paradigms that focus on either
single-species (e.g. indicators, umbrellas or flagships) or whole ecosystems
(ecological processes and habitats). Not surprisingly, both have their advocates and
detractors. The keystone species concept, which retains a species focus while
avoiding the need to examine every species, and emphasises processes that directly
(e.g. predation, competition) rather than indirectly (e.g. nutrient cycling) control
biodiversity, may allow managers to combine the best features of both these
paradigms. By itself, however, the concept is unlikely to provide a panacea for
biodiversity managers.
Keywords: Community importance index; context-dependent; functional role;
organism controlling potential dominant; resource provider; mutualist;
ecosystem engineer.
© July 2002, Department of Conservation. This paper may be cited as:
Payton, I.J.; Fenner, M.; Lee, W.G. 2002: Keystone species: the concept and its relevance for
conservation management in New Zealand. Science for Conservation 203. 29 p.
Science for Conservation 203
5
1. Introduction
The ‘keystone species’ concept has proved both promising and elusive in
theoretical and applied ecology. The term originates with Paine’s work on a
rocky shore community in California (Paine 1966). When he removed the top
predator (a starfish, Pisaster ochraceus) from a section of the shore, the
original 15-species assemblage was reduced to eight species. The collapse of a
system as a result of the removal of one of its elements prompted the architectural analogy with the keystone of an arch. Subsequently, numerous species
in a wide range of communities throughout the world have been claimed to
have keystone status, often with little experimental or quantitative evidence.
Power et al. (1996) dismiss many of these as ‘anecdotal keystones’.
The idea that some species may function as keystones has not been without its
critics (Mills et al. 1993; Hurlbert 1997), who argue that while the concept has
‘developed tremendous currency and fashionability’ it has also become increasingly ill-defined, and now means little more than ‘important for something’
(Hurlbert 1997). In defence of this position, opponents cite the vagueness of
keystone definitions and their inconsistent usage in the literature, and an
implied dichotomy between keystone and non-keystone species that has not
been demonstrated in nature. As the causal metaphor it was originally intended
to be (Paine 1995), Hurlbert concedes that the notion was ‘appealing and
harmless’, but as a well-defined concept or phenomenon he concludes it ‘has
had a stultifying effect on ecological thought and argument’. But if keystone
species do exist, and if they are widespread in nature, then the concept is
clearly of value in understanding community function, and has important implications for managing ecosystems to maintain or enhance their biodiversity (de
Maynadier & Hunter 1994; Simberloff 1998).
2. Keystone concepts
Of the various definitions of keystone species that have been attempted, perhaps
the most useful is that given by Power et al. (1996) as ‘a species whose effect is
large, and disproportionately large relative to its abundance’. It is not enough for a
species to be highly influential; its role must be great in relation to its relative
biomass contribution. Beech trees in a beech forest are not considered keystones
because their effect (while great) is not disproportionate to their relative
abundance. In Power’s terminology, such species are ‘dominants’. In the
architectural analogy, they might be considered ‘foundation stone’ species.
It is clear from the above definition that both aspects of keystones are a matter
of degree. How great an effect is ‘large’? How large a proportion is
‘disproportionate’? Power et al. (1996) suggest using the ‘community
importance’ (C.I.) index developed by Mills et al. (1993). This is the strength of
the effect (or impact) of a species on a community (as measured by the change
6
Payton et al.—Keystone species: the concept and its relevance in New Zealand
in a community trait such as productivity, species richness, etc.) per unit
change in a measure of the abundance (usually biomass) of the species. The
effect may be positive or negative. The easiest way to determine the C.I. of a
species is to remove the species entirely and measure the resulting change in a
relevant community feature. A species with an impact proportional to its
abundance would have a C.I. of 1. That of a keystone species would be greater
than unity, though how much greater is still not defined (Power & Mills 1995;
Power et al. 1996).
It seems probable that in any community the removal of individual species
would reveal a spectrum of types from so-called ‘redundant species’ (whose
removal has a negligible effect) to ‘keystone species’ (whose removal causes
major changes). The existence of redundant species may be implicit in the
concept of keystones because it suggests a hierarchy of species importance.
However, the idea of a redundant species has been criticised conceptually and
methodologically as a deficient paradigm. Gitay et al. (1996) point out the
difficulty of defining, detecting or demonstrating redundancy. Navarrete &
Menge (1996) show that previously ‘redundant’ species (whelks) with a low
‘interaction strength’ (the effect that one species exerts on another) can play a
major role in a community if the former keystones (starfish) are removed. Both
of the above papers argue strongly that, for conservation management
purposes, no species should be assumed to be redundant.
An important point to emerge from several studies is that the keystone status of
a species is entirely context-dependent. The community importance of the
same species may vary from one ecological situation to another. Menge et al.
(1994) demonstrated this in a series of experiments involving the original
keystone species, the starfish Pisaster ochraceus. They showed that this species
acts as a keystone by its predation of mussels (Mytilus californianus) on waveexposed rocky headlands but not in sheltered areas nearby. Navarrete &
Menge’s (1996) experiments similarly showed that whelks (Nucella spp.) acted
as keystones in the absence of starfish, but not in their presence. Even in
apparently similar contexts, the keystone role can be adopted by different
organisms. Fauth (1999) found that in comparable habitats in North and South
Carolina the keystone role was adopted by different amphibians. In this sense
‘keystoneness’ is not an intrinsic property of a species, but of its functional role
in a particular community assemblage. The diversity of organisms that have
been shown to act as keystones indicates that they cannot be readily predicted
on the basis of particular life-history traits. They can only be identified by
examining their function in specific communities. For example, the prairie dog
(Cynomys ludovicianus) has generally been considered a keystone species in
North American grasslands, but this species inhabits a wide variety of grassland
types. Stapp (1998) questions whether it necessarily plays a keystone role in all
of them. Each grassland type is a different ‘context’ requiring a separate
assessment of the prairie dog’s role.
If the removal of a single species can cause major changes to the structure,
function or diversity of a community, it follows that keystone-dependent
communities are perhaps more vulnerable and potentially unstable than
communities where keystones are absent. Power et al. (1996) point out that the
extinction of species may result in a community becoming keystone-dependent.
If several species share a trophic level, the removal of one of them may simply
Science for Conservation 203
7
result in compensatory action by the others as they expand in abundance or
niche breadth. However, if all but one of the species in the group are eliminated
(e.g. by human influence), the remaining species may come to play a keystone
role it did not previously have. The avian fauna of New Zealand, for example,
has been much depleted by human influence. It is possible that this may have
placed some of the surviving species (e.g. kereru Hemiphaga novaeseelandiae) in keystone roles they did not formerly have, or at least increased their
‘community importance’.
Another important way in which the ecological context can change markedly is by
the addition of alien species. An added species could become a new keystone in a
community where no equivalent or comparable species previously existed.
It has been suggested that rather than restricting the keystone concept to
individual species, it should also be applied to groups of organisms (Willson &
Halupka 1995) and major abiotic (e.g. frost, fire) processes (Bond 1994a).
However, this dilutes the original concept to a point where all communities
would have to be considered keystone-dependent. For example, every
functional group (grazers, predators, pollinators, seed dispersers, detritivores)
plays an essential role within its community, and the removal of even a subgroup would clearly be disruptive at some scale. Similarly, all communities are
influenced by major abiotic processes. By contrast, a keystone is, by definition,
a single object on which the forces holding an arch together are concentrated.
Only cases where the responsibility for keeping a community intact (or dramatically altering it) resides in one species can be truly considered keystones.
3. Types of keystone species
The keystone role for a species can arise in several different ways. Here we
distinguish four types of organism on the basis of their functional role as
keystones: (i) organisms controlling potential dominants; (ii) resource
providers; (iii) mutualists; and (iv) ecosystem engineers. These groups are not
mutually exclusive, and individual keystone species may exhibit characteristics
of more than one functional type.
3.1
ORGANISMS CONTROLLING POTENTIAL
DOMINANTS
The classic keystone is an organism that controls the population of a potentially
dominating species, thereby promoting coexistence by reducing competition
amongst other species for a limiting resource. In the case of Paine’s rocky shore
community (Paine 1966; 1974), the primary resource is space, the potential
dominant that tends to occupy all of the rock surface is the mussel (Mytilus
californianus), and the keystone species that keeps the mussels in check is the
starfish (Pisaster ochraceus). The removal of the mussels opens up enough
rock surface to prevent competitive exclusion eliminating a range of less
8
Payton et al.—Keystone species: the concept and its relevance in New Zealand
competitive species. Another example of a top predator controlling species
diversity in lower trophic levels is that of the triton gastropod Charonia, which
eats the starfish Acanthaster planci, thereby preventing over-consumption of
coral by the latter (Paine 1969). The removal of the triton results in loss of coral
and consequent community degradation.
In terrestrial ecosystems, herbivores play a major role in maintaining the
structure and species composition of the vegetation. The keystone status of
many of them can be readily determined by removal experiments that follow
the consequences for other elements of the community (e.g. Brown & Heske
1990). The effect of herbivores is often to suppress potentially dominant plant
species, reducing competitive exclusion of less competitive species, and
promoting a more species-diverse vegetation. Devries (1995) recognises the
central role of large herbivores in the maintenance of many European vegetation types and advocates their use in large-scale nature reserves. However,
large grazing herd-forming mammals might not qualify as keystones as their
effects might not be considered disproportionate to their biomass contribution
to the community. In contrast, the effect of some small herbivores may give
them keystone status. Experimental removal of kangaroo rats (Dipodomys spp.)
in the Chihuahua Desert profoundly altered the whole community (Brown &
Heske 1990; Heske et al. 1993; Kerley et al. 1997). Even very small herbivores
can exert a very marked effect. Hanley et al. (1995) found that slugs and snails
in grassland had a large influence on species composition and diversity because
their selective grazing on seedlings resulted in differential recruitment. The biomass of vegetation eaten was negligible, but the qualitative effect of its removal
was very marked, possibly giving the molluscs a keystone role in the community.
Grazers can act as keystones, but are often themselves top-down controlled by
predators, which are then considered to have the keystone role. The recently
reported case of sea otters (Enhydra lutris) on Adah Island, Alaska (Estes et al.
1998), provides an interesting example of the difficulties in deciding exactly
which species in a trophic sequence is the keystone. Sea otters are part of the
food chain: kelp forest, sea urchin (Strongylocentrotus spp.), sea otter, killer
whale (Orcinus orca). The recent increase in predation by killer whales on the
sea otters has resulted in a sharp decline in numbers of the latter species,
leading to an 8-fold increase in sea urchins, and a 12-fold decline in kelp density.
Loss of the kelp beds alters wave action and siltation rates with further effects
on other inshore species, such as barnacles (Balanus spp.) (Kaiser 1998). This
sequence of interactions has apparently been caused by killer whales shifting
their diet to sea otters because of the decline in their normal prey of seals (e.g.
Phoca vitulina) and sea lions (e.g. Eumetopias jubatus), itself due to the
decline in fish stocks in the Bering Sea because of overfishing. In such a lengthy
sequence of predators it is difficult to identify which species should be considered the keystone: one could as readily ascribe the role to the sea urchin, the
sea otter, the killer whale, or the human (Homo sapiens). The removal of any
one of these has a ‘domino’ effect on all the other organisms in the food chain.
This example illustrates the cascade principle first proposed by Darwin in the
Origin of Species in which he notes that the activities of cats (Felis catus) could
affect the population of red clover (Trifolium pratense). This was because cats
eat field mice (Apodemus sylvaticus), which damage the nests of humble bees
{bumble bees} (Bombus sp.) which, in turn, pollinate the clover flowers
Science for Conservation 203
9
(Darwin 1859, Ch.3). Keystone effects of this type are probably most frequent
where relatively simple food chains exist without strong collateral connections
that could absorb the removal of a particular link in the chain. Carpenter et al.
(1985) show how trophic interactions cascade in a lake community. Strong
(1992) suggests that cascades may be often aquatic with algae at the base; but
that in complex food webs, consumption in any particular trophic level is so
differentiated that the effect of removing any one species is normally buffered
by its functionally equivalent colleagues. He argues that true trophic cascades
imply keystone species, and that since all trophic interactions do not cascade,
simple top-down control may not be the norm.
Disease-producing organisms can also function as keystones, where their
impact on predator or herbivore populations has significant flow-on effects for
dominant elements of the wider community. Some of the best-known examples
are the rinderpest virus and the anthrax bacterium (Bacillus anthracis), which
have periodically devastated grazing mammal populations in parts of eastern
and southern Africa. Here the disease organism, by killing the herbivores and
reducing the grazing pressure, facilitated a pulse of woody species regeneration
and the displacement of extensive grassland communities by Acacia spp.dominated woodlands (Prins & van der Jeugd 1993). Fungal pathogens may
similarly fulfil a keystone role, most notably when introduced to communities
dominated by non-resistant host species. Well-documented examples include
chestnut blight (Cryphonectria parasitica), which has reduced the oncedominant American chestnut (Castanea dentata) to a minor understorey
component in eastern United States forests (Anagnostakis 1995), and Dutch elm
disease (Ophiostoma ulmi and O. nova-ulmi), which has similarly decimated
American elm (Ulmus americana) populations (Karnosky 1979). In both cases
the keystone role of the fungus is in removing a dominant canopy species,
which can be expected to have significant flow-on effects for the rest of the
forest community.
Keystone species have not been found in all communities where they have been
sought. An attempt by Wright et al. (1994) to test the hypothesis that tropical
forest regeneration was affected by the extinction of large carnivores through a
cascade effect gave ambiguous results. Tanner et al. (1994) found that high
species diversity in coral assemblages in the Great Barrier Reef was maintained,
not by a single keystone species, but by the action of physical disturbance
creating space for recruitment. One suspects that cases where keystones have
been sought but not found are less likely to be published than cases that
demonstrate their presence.
3.2
RESOURCE PROVIDERS
A second type of keystone species is one that provides a vital resource to a
range of organisms at a time of scarcity (which often occurs seasonally). The
resource provider itself may not be a particularly abundant species, but if it is
removed, the dependent organisms are unable to bridge the gap in supplies.
The most familiar examples involve the provision of keystone plant resources
such as fruit during seasonal shortages (Terborgh 1986a; Van Schaik et al.
1993). Peres (1994) found that a small number of plant species provided a range
10
Payton et al.—Keystone species: the concept and its relevance in New Zealand
of resources in the dry season that acted as temporary alternative sources of
nourishment for four primates in western Brazil. The resources included exudates,
nectar, immature seeds as well as fruit.
Many tropical forest communities include numerous obligate frugivores, mainly
birds, primates, bats and specialised insects. However, the availability of fruit
varies markedly through the year, with each plant species exhibiting its own
phenological pattern (Fenner 1998). ‘Bottle necks’ or ‘pinch points’ occur in
the supply of fruit, and the few species of plant that fruit at those times provide
a life-line that tides the frugivores over the period of dearth. Terborgh (1986b)
estimates that in the floodplain forest of Cocha Cashu in Peru, only a dozen
species of plants (out of trees, stranglers, vines and palms) in a total flora of
1000 species produce such keystone resources. In fact, it is possible that the
presence of keystone fruit providers may determine the species diversity of the
fruit-eaters. The Urucu terra firme forest in Brazil has a notably low density of
frugivores in comparison with equivalent forests elsewhere. Peres (1994)
suggests that this may be due to the severe periodic shortages of fruit that occur
because of the strongly seasonal production of fruit in this community.
The fig genus (Ficus) stands out as being of central importance as a keystone
resource provider in many tropical forests. One reason is the continuous
aseasonal fruiting in this genus, which may be a consequence of the fig’s
reliance on specialised wasp pollinators (Milton 1991). In Malaysian lowland
rainforests, at least 60 species of birds and 17 species of mammals eat figs.
Lambert & Marshall (1991) show that a guild of 29 Ficus species in a Malaysian
dipterocarp forest acted as keystone resource providers for a wide range of
frugivores. However, in some forests, e.g. in the Western Ghats in India, Ficus
trees are somewhat patchy in their distribution, and only act as keystone
resource providers for highly mobile species with large home ranges. For less
mobile animals, such as the Malabar giant squirrel (Ratufa indica), the figs
acted as keystones only for those individuals whose home range included fig
trees (Borges 1993). In other tropical forests, e.g. those in Gabon, figs are too
scarce to act as lean-period keystone providers, and the role is filled by two
species of Myristicaceae and one of Annonaceae for a range of monkeys and
frugivorous birds (Gautier-Hion & Michaloud 1989).
Microbial organisms may also function as keystone resource providers where
their presence enhances a limiting (usually nutritional) resource. On the island
of Hawaii the invasive symbiotic nitrogen-fixer (Myrica faya) acts as a keystone
resource provider in nitrogen-limited forest communities developing after
vulcanism (Vitousek 1990). The increased nitrogen availability has flow-on
effects for litter accumulation, earthworm populations (Aplet 1990) and
nutrient cycling, causing major changes to the formerly open-canopied forests,
including the displacement of dominant native tree species.
3.3
MUTUALISTS
Where two species are mutually dependent, the elimination of one will result in
the demise of the other. In this sense they act as keystones for each other. What
might be called ‘group mutualisms’ are common (e.g. between fruit producers
Science for Conservation 203
11
and seed-dispersing frugivores collectively, or between pollinators and groups
of plants). But one-to-one codependency is a risky strategy and is probably rare
in most communities. It is possibly more common in some types of organisms,
such as orchids and their specialist pollinators. Darwin (1862) gives some
remarkable examples of plant-pollinator co-evolution. Bond (1994b) describes
cases of pollination mutualisms in the fynbos vegetation of the Cape, South
Africa, where a number of plants are exclusively pollinated by a sparse fauna of
long-tongued flies, oil-collecting beetles, monkey beetles, carpenter bees and
butterflies. Certain birds in Malaysia are fig-eating specialists (Lambert &
Marshall 1991), but the mutualism here is probably unequal as the fig does not
depend exclusively on these specialist feeders for seed dispersal. More
convincing cases of mutualists acting as true keystones arise where there is
dependency of several species on a single mutualist. Manning & Goldblatt
(1997) give an example of a guild of at least eight (unrelated) plant species with
long tubular flowers, all of which depend exclusively on a single species of
long-proboscid fly for pollination. Clearly, the insect here fulfils a keystone
role.
Even a one-to-one mutualist could act in a keystone capacity if at least one of the
partners had strong interactions with other species in the community. The
removal of the mutualist pollinator of a dominant plant would markedly disrupt
the community. Fig-wasps provide an interesting example of this. Ficus trees
and fig-wasps are mutually dependent. The trees act both as major structural
elements in the community and as keystone resource providers for frugivores.
These frugivores are, in turn, essential seed dispersal agents for a wide range of
other plants (Anstett 1997). The fig-wasp has an influence greatly disproportionate to its abundance, and is perhaps one of the most extreme examples
of a keystone mutualist. The vulnerability of this mutualism is suggested by the
recent work of Nason et al. (1998) showing that Ficus trees in the small Barro
Colorado Island nature reserve, Panama, depend on cross-pollination from trees
distributed over a wide area of the surrounding forest. The reserve is thus not
self-contained for this keystone resource provider as it depends on forest
management beyond its borders.
3.4
ECOSYSTEM ENGINEERS
Organisms can also act as keystones by modifying the physical environment in
ways that release resources for other species. The activities of many organisms
provide habitats that would not otherwise be available, often by means of
disturbance to the physical habitat. Because of the structural alterations they
bring about, such organisms have been referred to as ecosystem engineers
(Jones et al. 1994). They can be considered keystone species if their effects are
large and disproportionate to their abundance.
Examples of the action of ecosystem engineers are legion. They include dam
building by beavers, which has major local effects on both terrestrial and aquatic
habitats (Naiman et al. 1988); soil disturbance as a result of burrowing by a wide
range of species such as rabbits, moles, badgers, prairie dogs, gophers and
kangaroo rats; nest building by ants and termites; wallow creation by alligators,
12
Payton et al.—Keystone species: the concept and its relevance in New Zealand
vegetation destruction by elephants, and hole drilling by woodpeckers. The longterm engineering effects of earthworm action in soil have been known for over a
century (Darwin 1881).
In most of these cases the provision of habitat or other resources is an
incidental by-product of the engineer’s activity. The benefiting organisms can
be considered opportunists that have become adapted to exploit the resource.
A good example of an ‘incidental’ provider of both habitat and food is the rednaped sapsucker (Sphyrapicus nuchalis) of the Colorado subalpine meadows
(Daily et al. 1993). Its abandoned nesting holes in aspens are occupied by two
species of swallow, whilst its feeding holes in willows provide sap flows for a
range of birds, mammals and insects. Its engineering activities are thus quite
modest, but the results are of disproportionate importance.
In some cases there is a high degree of adaptation in the benefiting species.
Redford (1984) recorded the inquiline organisms (non-parasitic occupants) in
termite mounds (of Cornitermes cumulans) in the Brazilian cerrado community. These included at least 17 other species of termite and ten species of
ants (many of them obligate inquilines), as well as many other insects (beetles,
flies, wasps and bees) and other invertebrates (scorpions, centipedes, spiders)
and vertebrates (lizards, snakes, mice and birds). Termites are true ecosystem
engineers and can be considered to have a keystone status in this region as their
elimination would have a major effect on community structure and species
diversity.
In assessing the keystone role of an organism it is not always possible to
distinguish clearly between the effects of predation and engineering. A good
case can be made for considering large grazers as keystone ecosystem
engineers, as well as herbivores. By suppressing invasion of woody species and
maintaining open grassland vegetation they profoundly alter the physical
structure of the community. Even the classic keystone predator, the starfish
Pisaster, exerts its influence not just as a predator but as the incidental creator
of bare patches, which provide recruitment sites for other species. Most
organisms probably alter their environment in some way that can be exploited
by other species. Whether the engineer can be considered a keystone species is
a matter of degree: how profound the physical effect is, how disproportionate it
is, how many other species benefit, and how great their dependence on the
alteration is.
Science for Conservation 203
13
4. The New Zealand context
In the late 1960s Paine visited New Zealand and repeated his original predator
removal experiment (Paine 1966) on the rocky intertidal communities at
Anawhata, just north of Piha, near Auckland (Paine 1971; Paine et al. 1985).
During the 9-month period that the top carnivore (a starfish, Stichaster
australis) was removed from the experimental plots, the area occupied by its
primary prey (the green-lipped mussel, Perna canaliculus) almost trebled, the
mussel expanded its intertidal range, and the species assemblage of the
experimental plots reduced from 20 to 14. Since that time there have been few
references to keystone species, or studies of their functional importance, in the
New Zealand ecological literature. There are, however, numerous examples of
organisms that may fulfil a keystone role in our natural ecosystems.
4.1
ORGANISMS CONTROLLING POTENTIAL
DOMINANTS
The decline of much of the native fauna has been linked to introduced predators
such as feral cats, rats (including kiore) and stoats (Holdaway 1999; Innes et al.
1999; Lloyd & Powlesland 1994). However, for the predators to be considered
keystones the removal of the prey species needs to affect dominant components of the ecosystem, and to date few such links have been clearly demonstrated. In freshwater ecosystems, exotic fish species such as brown trout
(Salmo trutta) and rainbow trout (Oncorhynchus mykiss) are credited with the
decline of native fish (Galaxias spp. and Gobiomorphus spp.), crayfish
(Paranephrops spp.) and insect populations (Cadwallader 1975; McDowall
1968, 1987; Townsend & Crowl 1991), and may be a factor in the decline of the
blue duck (Hymenolaimus malacorhynchos) (Towers 1996). Trout species
may warrant keystone status where their presence has a disproportionate effect
on the abundance or biomass of the freshwater biota. However, because the
impact of adjoining land use tends to dominate stream ecology, demonstrating a
keystone role for trout may not be a straightforward process.
The keystone role of introduced herbivore species is more readily demonstrated. On Stewart Island, where coastal forests are periodically subject to saltspray-related dieback, white-tailed deer (Odocoileus virginianus), by preventing the regeneration of the canopy tree species, play a keystone role in the
replacement of these forests by shrubland, fernland or grassland communities
(Stewart & Burrows 1989). A similar situation occurs after fire in podocarp–
tawa (Beilschmiedia tawa) forests in the northern Urewera ranges (Payton et
al. 1984). Where previously, dense pole stands of kanuka (Kunzea ericoides)
acted as a nurse crop, initially for kamahi (Weinmannia racemosa) and later for
the podocarp/hardwood species characteristic of the pre-fire forests, deer now
prevent kamahi regeneration and maintain the dominance of the early
successional kanuka tree species. Australian brushtail possums (Trichosurus
vulpecula), by defoliating and killing canopy trees, can also be considered to
play a keystone role in many New Zealand forests. Where the dominant tree
14
Payton et al.—Keystone species: the concept and its relevance in New Zealand
species are preferred foods, possum browsing may lead to the collapse of forest
canopies over large areas (Batcheler 1983; Rose et al. 1992), and their
replacement by forest or shrubland communities less palatable to the ungulate
browsers. In both cases the keystone role of the herbivore results from the large
and disproportionate influence it has on the dominant elements of the forest
community, which is exacerbated by the absence (except for humans) of topdown predator control. In the short term (< 150 years since the introduction of
mammalian herbivores) the keystone function of the ground browsers is most
evident where forests are regenerating after disturbance, but over the lifetime
of the forest both ground and arboreal browsers can be expected to have an
overriding influence on the composition and structure of the vegetation.
Vertebrate herbivory does not invariably result in the removal of native
dominants. In some communities rabbit (Oryctolagus cuniculus) and sheep
(Ovis aries) grazing fulfils a positive keystone role by preventing dominant
native plant species being overtaken by more aggressive adventive grasses and
other weeds. The cessation of sheep grazing significantly reduced the number
of both indigenous and adventive plant species, and increased the dominance of
adventive grasses such as cocksfoot (Dactylis glomerata) in highly modified
silver tussock (Poa cita) grasslands on the Port Hills near Christchurch (Lord
1990). On Motunau Island off the north Canterbury coast, the removal of rabbits
(Cox et al. 1967) resulted in much of the silver tussock grassland that formerly
dominated the island being replaced by adventive shrubs, forbs and grasses
(I. Payton, pers. obs.), which were previously suppressed by the rabbits.
Similarly, exclusion of sheep from Danthonia (Rytidosperma spp.) grassland
reserves on the Canterbury Plains has allowed the browse-tolerant native
grasses to be displaced by more aggressive adventive grasses (Meurk et al.
1989). In each case the selective grazing pressure of the keystone herbivore is
required to maintain the indigenous character and/or the diversity of the
grasslands.
Rabbit haemorrhagic disease (RHD), which was illegally introduced into New
Zealand in August 1997, may also have a keystone role in at least some grassland
ecosystems. In addition to changes to vegetation composition resulting from
fewer rabbits and reduced levels of herbivory (Norbury & Norbury 1996), prey
switching by rabbit predators is expected to adversely affect native grassland
fauna, at least in the short term (Norbury 1999). The likely keystone role for the
virus results from its ability to initiate disproportionately large changes in plant
and animal biomass and diversity within the grassland, relative to its own
abundance. The recently isolated disease organism ‘Phytoplasma australiense’,
which causes flax yellow leaf, cabbage tree sudden decline (Andersen et al.
1998a, b) and coprosma lethal decline (R. Beever, pers. comm.) may also
warrant keystone status, at least in plant communities where the affected plant
species predominate.
The keystone status of each of these organisms is clearly context-dependent.
Although the possum may function as a keystone species in rata–kamahi forests, the
same cannot be said for its role in beech-dominated ecosystems where floristic
composition but not forest structure is typically affected (Wardle 1984). Rabbit
haemorrhagic disease has significantly reduced rabbit populations in many semiarid grassland communities, but has thus far had little effect in areas of higher
rainfall and would not be expected to play a keystone role in these ecosystems.
Science for Conservation 203
15
Likewise the cabbage tree sudden decline epidemic, which began in the early
1980s, has resulted in the loss of large numbers of cabbage trees in the northern
half of the North Island, but has had little or no effect on South Island cabbage
tree populations (Beever et al. 1996).
Invasive plant species are also potential keystones where they displace
dominant elements and substantially alter the species composition of existing
communities. In remnant and disturbed forest stands Old Man’s Beard (Clematis
vitalba) vines alter forest structures by smothering mature trees (Williams &
Timmins 1990; West 1992) and, together with the adventive herb Wandering
Willie (Tradescantia fluminensis), affect plant species composition by
inhibiting seedling regeneration (Kelly & Skipworth 1984). On alluvial sites the
naturalised shrub buddleia (Buddleja davidii) rapidly displaces primary native
colonisers, and can accelerate successions to mature forest by replacing longerlived natives such as kanuka (Smale 1990). Adventive hawkweed (Hieracium)
species play a similar role in many grassland communities, displacing the
dominant tussocks (Makepeace 1985) and reducing indigenous species diversity
(Rose et al. 1995).
In freshwater ecosystems such as Lake Taupo, the adventive macrophyte
Lagarosiphon major has significantly altered plant species composition and
biomass over large areas of the littoral zone. In addition to directly displacing
the native plant species, Lagarosiphon acts as a keystone resource provider for
adventive black swan (Cygnus atratus) populations (Bull 1983) whose grazing
activities also uproot native macrophyte (e.g. Potamogeton spp.) communities
(Howard-Williams & Davies 1988). Increased detritus resulting from the grazing
activities of the swans provides an added food source for native koura
(freshwater crayfish, Paranephrops planifrons) populations, which in their
turn influence species richness and invertebrate biomass (Parkyn et al. 1997),
and limit the depth to which native macrophyte communities extend (Coffey &
Clayton 1988).
4.2
RESOURCE PROVIDERS
Organisms that may act as keystone resource providers can be found at a range
of trophic levels in New Zealand ecosystems. The low levels of nitrogen and
phosphorus in many New Zealand soils provide a potential keystone role for
nitrogen-fixing organisms and mycorrhizal fungi. Nitrogen-fixing bacteria are
known from native plant genera such as Carmichaelia, Coriaria, Discaria, and
Sophora (Greenwood 1978; Torrey 1978) that are, typically, early colonisers of
disturbed sites, and are the basis for legume introductions to enhance grassland
productivity (Sears 1960) and assist the establishment of exotic pine plantations
on sandy soils (Gadgil 1971). Where phosphorus is the nutrient limiting plant
growth, native mycorrhizal fungi have been shown to be essential for the
growth and survival of a wide range of native tree and shrub species.
Conversely, the lack of suitable mycorrhizal fungi has been used to explain the
slow spread of beech (Nothofagus) species into adjacent plant communities
(Baylis 1980). In both cases the possible keystone role for the resource provider
results from changes to plant community composition initiated by the increased
nutrient availability. Both examples can also be viewed as keystone organisms
16
Payton et al.—Keystone species: the concept and its relevance in New Zealand
controlling potential dominants and, where the micro-organism:host plant
relationship is close and/or obligate, as keystone mutualists.
For island ecosystems Daugherty et al. (1990) have suggested a keystone role
for seabird populations whose guano, by enriching soil fertility, supports high
densities and diversities of invertebrates and reptiles. However, closer examination of the influence of seabirds and rats in island ecosystems (Markwell
1999) has concluded that although both groups have important roles on
offshore islands, neither conforms to the definition of a keystone species.
Amongst the native invertebrates perhaps the clearest example of a keystone
resource provider are the beech scale insects, Ultracoelostoma spp., which
secrete a sugary exudate known as honeydew (Morales et al. 1988). In beech
forests where these insects are common, honeydew is an important food source
for nectar-feeding birds such as tui (Prosthemadera novaeseelandiae), bellbird
(Anthornis melanura) and silvereye (Zosterops lateralis) (Gaze & Clout 1983),
particularly in the winter (Wardle 1984). Over recent decades the availability of
this abundant high-energy food supply has allowed two invading species of
Vespula wasps to reach high densities in honeydew beech forests (Thomas et
al. 1990). In addition to directly affecting the nectar-feeding birds, wasp
predation has depleted invertebrate populations (Beggs & Rees 1999) which, in
turn, has adverse effects for insectivorous bird populations in these forests
(Moller & Tilley 1989; Moller et al. 1993). The keystone role for the scale insect
results both from the direct provision of a vital food source at times of scarcity,
and from the indirect consequences of this resource for other groups of
organisms within the beech forest ecosystem.
Obligate frugivores are uncommon in the New Zealand avifauna and elsewhere.
While a relatively high proportion of New Zealand’s native plants (c. 250
species) have fleshy fruits, fewer than 12 native bird species are regular
frugivores and only the endemic flightless parrot kakapo (Strigops habroptilus)
is thought to be dependent on periodic heavy (mast) fruiting years for breeding
success. Availability of fruit varies markedly between plant communities (e.g.
podocarp–hardwood forests > beech forests), seasons (late summer à late
winter > early spring à summer) and years (Clout & Hay 1989; Lee et al. 1990).
Fleshy-fruited species that may fulfil a keystone role for native frugivores fall
into three categories, (i) those that enable the build-up of energy reserves
before winter (e.g. kereru consumption of tawa and miro (Prumnopitys
ferruginea) berries—McEwan 1978), (ii) those that provide a reliable source of
fruit during periods when alternative foods are scarce (e.g. supplejack
Ripogonum scandens), and (iii) those where periodic heavy (mast) fruiting is
associated with successful breeding (e.g. kakapo consumption of pink pine
Halocarpis biformis seed—Powlesland et al. 1992).
For the takahe (Porphyrio mantelli), a flightless endemic rail now restricted to the
Murchison Mountains in Fiordland, rhizomes of the summer-green fern Hypolepis
millefolium provide a critical energy source during winter when snow forces birds
from their summer grassland range (Mills et al. 1980). Similarly, high levels of
available carbohydrate in fern (Blechnum spp.) and clubmoss (Lycopodium
ramulosum) rhizomes provide a reliable source of energy for the kakapo (James et
al. 1991). In both cases the possible keystone role of the plants is the provision of
an essential resource in an environment and at a time of the year when alternative
sources are scarce.
Science for Conservation 203
17
4.3
MUTUALISTS
Specialist pollinators are not a feature of the New Zealand flora (Lloyd 1985),
and there are few documented examples of one-to-one codependency (Webb &
Kelly 1993). The recently described cases of bat pollination of the root parasite
Dactylanthus (Ecroyd 1996), and honeyeater pollination of mistletoe flowers
(Ladley & Kelly 1995; Ladley et al. 1997) may warrant consideration as potential
keystone relationships. Video camera surveillance of Dactylanthus flowers
suggests this species is pollinated almost exclusively by an endangered endemic
bat (Mystacina tuberculata), but that in the absence of the bats the introduced
ship rat (Rattus rattus) can assume this role. For the Peraxilla mistletoes,
pollination is dependent on two species of endemic honeyeater (tui and
bellbird), which twist the top of the ripe buds and trigger an explosive opening
mechanism. Flowers not visited by birds fail to open and only a minority set
seed. In both cases seed production and hence the long-term survival of the
plant species is heavily, but not entirely, dependent on a single pollinator or
pollinator-group. In neither case is the affected plant species either a dominant
member of its community or a demonstrated part of a wider trophic cascade
affecting dominant elements of a community.
The decline or extinction of much of the native terrestrial avifauna raises
questions about their role as seed dispersers in pre-human times, whether
guilds of native seed dispersers may now be absent or reduced to single
(perhaps keystone) species, and the extent to which introduced species now
fill these roles. Among the extinct avifauna, moas (Dinornithiformes), huia
(Heteralocha acutirostis) and piopio (Turnagra capensis) ate fruit or seeds of a
wide range of tree and shrub species (Burrows 1989; Clout & Hay 1989), but
their demise has not resulted in the obvious regeneration failure of any native
plant species (Lee et al. 1990). Of the extant frugivores, Clout & Hay (1989)
consider kereru the most important seed disperser in New Zealand forests
because of its catholic diet (≥ 70 plant species), mobility, and widespread
distribution. The kereru may also warrant keystone status as the sole surviving
disperser of larger (>10–12 mm) seeded native tree species (Lee et al. 1990);
although Webb & Kelly (1993) consider this may overstate the importance and
effectiveness of seed dispersal by birds. For the native loranthaceous mistletoes,
where germination is entirely dependent on bird dispersal to remove the
exocarp, Ladley & Kelly (1996) conclude that while the current size of native
frugivore (bellbird, silvereye, tui) populations does not appear to threaten
mistletoe survival, the [keystone] role of the dispersers needs to be considered
when assessing historical or future mistletoe declines. Reductions in native
frugivore populations have been partly offset by introduced European
frugivores such as the blackbird (Turdus merula) and song thrush
(T. philomelos) (Clout & Hay 1989). In addition to dispersing many of the same
fruits as their native counterparts, the introduced frugivores target a wider
variety of adventive weed species, and may play a keystone role in the changing
composition of secondary shrub and forest communities (Burrows 1994;
Williams & Karl 1996). The extent to which introduced brushtail possums,
which also consume a wide variety of fruits (Cowan 1990), may fill the seed
dispersal role formerly held by native avian frugivores may also be worth
considering.
18
Payton et al.—Keystone species: the concept and its relevance in New Zealand
The possibility that at least some native lizards act as keystone mutualists also
warrants examination. Geckos (Hoplodactylus spp.) and skinks (Oligosoma
and Cyclodina spp.) include considerable quantities of fruit in their diets, and
while foraging distances are small (cf. frugivorous birds) this may be partially
offset by the longer time that seeds take to pass through the gut (Whitaker
1987). If keystone mutualist relationships involving lizards exist, they are most
likely to be found among divaricate shrubs such as Melicytus alpinus, where
fruits are inaccessible to many birds. Historically, the now extinct
Hoplodactylus delcourti (the largest known gecko) may also have played a
[keystone] role in the dispersal of large fruits (Webb & Kelly 1993).
4.4
ECOSYSTEM ENGINEERS
As with other keystone categories, some of the clearest New Zealand examples
of ecosystem engineers are found amongst introduced plant and animal species.
Until recent decades, introduced herbivores such as deer (e.g. Cervus spp.),
goats (Capra hircus), chamois (Rupicapra rupicapra) and thar (Hemitragus
jemlahicus) were seen as the primary cause of accelerated high-country erosion
(Holloway 1959). While natural erosion associated with geological instability is
now known to be more widespread than previously thought (Mosley 1978;
Grant 1989; McSaveney & Whitehouse 1989), there can be little doubt that the
impact of wild animal populations on understorey and litter layers has lessened
the ability of native forest and grassland communities to absorb and retain
moisture, and has altered hydrological regimes and biogeochemical cycles.
Adventive plant species also modify catchment hydrology. Willows (Salix spp.)
used to stabilise riverbanks and estuarine plantings of cord grass (Spartina
spp.) trap sediment, reduce the capacity of rivers, estuaries and wetlands to
accommodate flood events (Partridge 1987; Russell 1994), and threaten
biological diversity (Lee & Partridge 1983; Collier 1994). Replacement of native
(Chionochloa) grasslands with plantation (mainly Pinus spp.) forests reduces
both the magnitude of flood events and stream flows during periods of low
rainfall (Fahey & Jackson 1997) which, in turn, affects aquatic flora and fauna
and downstream ecosystems. The deeper-rooted conifers also increase nutrient
availability in the topsoil (Davis & Lang 1991), which can be expected to
provide a competitive advantage for adventive forbs and grasses with higher
nutrient requirements than their native counterparts. In the braided river
systems of the eastern South Island the spread of willows and Russell lupins
(Lupinus polyphyllus) has resulted in the channelling of formerly meandering
streams, a loss of open riverbed habitat for several rare or endangered birds
(e.g. black stilt Himantopus novaezelandiae, wrybill plover Anarchynchus
frontalis), and increased cover for their predators (Hughey & Warren 1997;
Maloney et al. 1999).
Other adventive plant species that may function as ecosystem engineers include
marram grass (Ammophila arenaria), gorse (Ulex europaeus) and Hakea spp.
Marram, whose roots and stems are more effective sand binders than the native
pingao (Desmoschoenus spiralis), changes the morphology and disturbance
regimes in sand dune systems (Esler 1978). Gorse and Hakea spp., which have
Science for Conservation 203
19
high rates of litter production and accumulation, have the potential to alter fire
regimes (Druce 1957; Fugler 1982) and affect the biota over large areas.
Amongst the introduced fauna, rabbits, whose burrows endanger seabird (Cox
et al. 1967) and sealion (Phocarctos hookeri) (Crawley 1990) populations, feral
pigs (Sus scrofa), whose rooting habits provide ideal establishment sites for
beech seedlings (Wardle 1984), and European or koi carp (Cyprinus carpio),
which cause significant damage to freshwater ecosystems by undermining stream
banks (McDowall 1990), may warrant consideration as ecosystem engineers.
5. Identifying keystone species
Identifying keystone species is problematic (Power et al. 1996), and attempts to
develop a set of species traits that would a priori determine keystone interactions
have thus far proved elusive (Menge et al. 1994). A variety of approaches have been
used, either singly or in combination, including experimental manipulations, comparative studies, natural history observations, and ‘natural experiments’ (Power et
al. 1996), but no robust methodologies have been developed.
Experimental removal is the most convincing way of demonstrating the
[keystone] role of an individual species in a particular community, but is
frequently impractical for logistical, technical, social, or ethical reasons. To
avoid the need for numerous manipulations to determine the presence and
identity of the keystone species in a community, Power et al. (1996) suggest
combining manipulative experiments with modelling approaches such as path
analysis (Wootton 1994). Using this approach, one or a few strongly interacting
species are experimentally reduced or removed, and the response to these
changes is monitored in a much wider range of species within the community.
Interpretation of removals is not always straightforward. Where experimental
treatments (e.g. exclosure fences) remove more than one species (e.g. rabbits
and sheep), it is often not possible to determine whether changes to the
unmanipulated state result from the actions of a single keystone species, or the
collective impacts of several members of a guild or trophic level (Power et al.
1996).
The time required to determine the consequences of experimental manipulations for other species in the community and the spatial scales over which the
results are applicable are both extremely variable (Power et al. 1996). As a
general rule, the impact of species manipulations occurs more rapidly in aquatic
ecosystems (Estes 1995). For example, while Paine (1971, 1974) was able to
demonstrate significant changes resulting from the removal of starfish from
rocky shore communities within less than 12 months, the full effects of
removing top predators from forest ecosystems may not become apparent for
decades or even centuries (Dirzo & Miranda 1990; Terborgh 1986b).
The use of natural history observations, or comparative studies which substitute space for time (i.e. use habitats with varying densities of putative keystone
species), overcomes many of the limitations of experimental manipulations. By
themselves both approaches lack the rigour of experiments, the former because
20
Payton et al.—Keystone species: the concept and its relevance in New Zealand
of its subjective nature and the latter because no two sites are ever exactly the
same. However, when used in conjunction with experimental manipulations,
both can be extremely useful for identifying hypotheses and determining the
generality of experimental results (Paine 1995; Power et al. 1996).
Large-scale changes in species abundance resulting from the invasion of
organisms into new environments, the overharvesting of natural populations,
or the deliberate eradication of pest species also offer opportunities for
determining the potential keystone role of species. The use of such ‘natural
experiments’ overcomes many of the social and ethical problems associated
with deliberate experimental manipulations. Problems frequently associated
with this approach include the lack of a pretreatment baseline, an inability to
find comparable non-manipulated sites, and the replication of treatments
(Carpenter 1989).
6. Implications for conservation
management
Conservation managers are faced with a bewildering array of species, communities and ecosystems, each subject to a variety of internal (e.g.
successional) and external (e.g. pest and weed) influences, which impose
continual change at all levels. Faced with this complexity and ever-increasing
demands for limited resources, the question still remains: what is the best way
to maintain indigenous biodiversity and retain functional natural ecosystems?
The inability to monitor and manage all aspects of biodiversity has led to the
development of paradigms that focus on either single-species (e.g. indicators,
umbrellas or flagships) or whole ecosystems (Tracy & Brussard 1994; Simberloff
1998). Not surprisingly, each have their advocates and detractors. Among the
single-species approaches, indicators create disagreement over what they
indicate (Landres et al. 1988), umbrellas (species requiring large areas of natural
habitat for their continued survival) raise questions over the range of biological
diversity they shelter (Franklin 1994), and flagships (iconic species) are often
notoriously expensive to keep afloat (Shrader-Frechette & McCoy 1993) and
may require management regimes that clash with other elements of the natural
biota (Ehrlich et al. 1992).
In contrast, ecosystem or landscape-scale paradigms emphasise ecological
processes (e.g. nutrient cycling) and habitats rather than individual species
(Franklin 1993; Meffe & Carroll 1994), using the rationale that biological
diversity is best preserved by maintaining healthy ecosystems (Bourgeron &
Jensen 1993). To do otherwise, proponents argue, will exhaust our available
time and financial resources, society’s patience, and scientific knowledge, long
before we make significant progress towards the preservation of existing
biodiversity (Franklin 1993). Opponents contend that where processes become
the focus of biodiversity management, the process rather than the species it
supports becomes what is valued. Because most ecosystem processes are
Science for Conservation 203
21
common to many species, the disappearance of threatened species may cause
little or no detectable change to any key process. Further, they argue that
whereas species are readily defined, ecosystems are arbitrary constructions that
can range from individual water droplets to the entire biosphere (Tracy &
Brussard 1994). The use of adaptive management regimes (i.e. ongoing modification of management practices based on observations of their outcomes) to
implement ecosystem-based approaches also comes in for criticism on the basis
that while they may provide ‘rules-of-thumb’ for managers, they tell us little
about the mechanisms that maintain species, communities or ecosystems
(Simberloff 1998).
In a paper that discusses options for managing biodiversity, Simberloff (1998)
suggests that the keystone species concept may allow conservation managers to
combine the best features of single-species and ecosystem-based management
approaches. He argues that the use of keystones retains a species focus while
avoiding the need to examine every species, and emphasises the mechanisms
that directly (e.g. predation, competition) rather than indirectly (e.g. nutrient
cycling) control biodiversity. Where it is possible to identify a keystone that is
critical for the continued survival (or demise) of many other species in its
community, management of that keystone may be an efficient means of
managing a much wider range of biodiversity.
By themselves, keystone species are unlikely to provide a panacea for biodiversity managers. Not all (natural) ecosystems of interest may contain
keystone species, and even where keystones are identified they may not be
easily managed as part of a conservation strategy. However, even if this worstcase scenario eventuates, the search will have provided us with much useful
information on the structure and functioning of ecosystems that will assist our
attempts to conserve biodiversity.
To date there have been few detailed studies of keystone species, and almost
none relating to New Zealand ecosystems. The challenge for New Zealand
researchers, whether or not they embrace the keystone species concept, is to
determine the interactions and functional significance of individual [keystone]
species and guilds within our natural ecosystems. This approach will allow us to
identify a hierarchy of importance amongst cohabiting species, and can be
expected to produce critical information for conservation efforts to preserve
indigenous biodiversity.
7. Acknowledgements
We thank Craig Millar (Department of Conservation, Hokitika) for initiating this
review, Gary Barker (Landcare Research, Hamilton), David Towns (Department
of Conservation, Auckland) and an anonymous referee for reviewing the
manuscript. The readability of the report was improved by Christine Bezar, and
Wendy Weller completed the final word processing.
22
Payton et al.—Keystone species: the concept and its relevance in New Zealand
8. References
Anagnostakis, S.L. 1995: The pathogens and pests of chestnuts. Advances in Botanical Research 21:
125–145.
Andersen, M.T.; Beever, R.E.; Gilman, A.C.; Liefting, L.W.; Balmori, E.; Beck, D.L.; Sutherland, P.W.;
Bryan, G.T.; Gardner, R.C.; Forster, R.L.S. 1998a: Detection of phormium yellow leaf
phytoplasma in New Zealand flax (Phormium tenax) using nested PCRs. Plant Pathology
47: 188–196.
Andersen, M.T.; Liefting, L.W.; Goldsmith, D.; Longmore, J.; Beck, D.L.; Beever, R.E.; Forster, R.L.S.
1998b: Detection of PYL phytoplasma in cabbage trees (Cordyline australis) affected with
sudden decline. Pp. 54–55 in Program and abstracts, 12th International Organisation of
Mycoplasmology Conference, Sydney, 22–28 July 1998.
Anstett, M.C.; Hossaert-McKey, M.; McKey, D. 1997: Modeling the persistence of small populations
of strongly interdependent species: Figs and fig wasps. Conservation Biology 11: 204–213.
Aplet, G.H. 1990: Alteration of earthworm community biomass by the alien Myrica faya in Hawai’i.
Oecologia 82: 414–416.
Batcheler, C.L. 1983: The possum and rata-kamahi dieback in New Zealand: a review. Pacific Science
37: 415–426.
Baylis, G.T.S. 1980: Mycorrhizas and the spread of beech. New Zealand Journal of Ecology 3: 151–153.
Beggs, J.R.; Rees, J.S. 1999: Restructuring of Lepidoptera communities by introduced Vespula wasps
in a New Zealand beech forest. Oecologia 119: 565–571.
Beever, R.E.; Forster, R.L.S.; Rees-George, J.; Robertson, G.I.; Wood, G.A.; Winks, C.J. 1996: Sudden
Decline of cabbage tree (Cordyline australis): search for the cause. New Zealand Journal of
Ecology 20: 53–68.
Bond, W.J. 1994a: Keystone species. Pp. 237–253 in Schultz, E-D.; Mooney, H.A. (Eds): Biodiversity
and ecosystem function. Springer, New York.
Bond, W.J. 1994b: Do mutualisms matter? Assessing the impact of pollinator and dispersor
disruption on plant extinction. Philosophical Transactions of the Royal Society of London
B 344: 83–90.
Borges, R.M. 1993: Figs, Malabar giant squirrels, and fruit shortages within two tropical Indian
forests. Biotropica 25: 183–190.
Bourgeron, P.S.; Jensen, M.E. 1993: An overview of ecological principles for ecosystem
management. Pp. 51–63 in Jensen, M.E.; Bourgeron, P.S. (Eds): Eastside forest ecosystem
health assessment, Vol. II. Ecosystem management: principles and applications. US
Department of Agriculture National Forest System, Washington, DC.
Brown, J.H.; Heske, E.J. 1990: Control of a desert-grassland transition by a keystone rodent guild.
Science 250: 1705–1707.
Bull, P.C. 1983: Aquatic birds of the lake and its surroundings. Pp. 133–150 in Forsyth, D.J.; HowardWilliams, C. (Eds): Lake Taupo. DSIR Science Information Publishing Centre, Wellington.
Burrows, C.J. 1989: Moa browsing: evidence from the Pyramid Valley mire. New Zealand Journal of
Ecology 12: 51–56.
Burrows, C.J. 1994: Fruit types and seed dispersal modes of woody plants in Ahuriri Summit Bush,
Port Hills, western Banks Peninsula, Canterbury, New Zealand. New Zealand Journal of
Botany 32: 169–181.
Cadwallader, P.L. 1975: Feeding relationships of galaxiids, bullies, eels and trout in a New Zealand
river. Australian Journal of Marine and Freshwater Resources 26: 299–316.
Carpenter, S.R. 1989: Replication and treatment strength in whole-lake experiments. Ecology 70:
453–463.
Science for Conservation 203
23
Carpenter, S.R.; Kitchell, J.F.; Hodgson, J.R. 1985: Cascading trophic interactions and lake
productivity. BioScience 35: 634–639.
Clout, M.N.; Hay, J.R. 1989: The importance of birds as browsers, pollinators and seed dispersers in
New Zealand forests. New Zealand Journal of Ecology 12 (supplement): 27–33.
Coffey, B.T.; Clayton, J.S. 1988: Contrasting deepwater macrophyte communities in two highly
transparent New Zealand lakes and their possible association with freshwater crayfish
(Paranephrops spp.). New Zealand Journal of Marine and Freshwater Research 22: 225–230.
Collier, K. 1994: Effects of willows on waterways and wetlands - an aquatic habitat perspective. Pp.
33–41 in West, C.J. (Comp.): Wild willows in New Zealand: proceedings of a Willow Control
Workshop hosted by Waikato Conservancy, Hamilton, 24–26 November 1993. Department
of Conservation, Wellington.
Cowan, P.E. 1990: Brushtail possum. Pp. 67–98 in King, C.M. (Ed.): The handbook of New Zealand
mammals. Oxford University Press, Auckland.
Cox, J.E.; Taylor, R.H.; Mason, R. 1967: Motunau Island, Canterbury, New Zealand. An ecological
survey. New Zealand DSIR Bulletin 178. 112 p.
Crawley, M.C. 1990: New Zealand sea lion. Pp. 256–262 in King, C.M. (Ed.): The handbook of New
Zealand mammals. Oxford University Press, Auckland.
Daily, G.C.; Ehrlich, P.R.; Haddad, N.M. 1993: Double keystone bird in a keystone species complex.
Proceedings of the National Academy of Sciences of the USA 90: 592–594.
Darwin, C.R. 1859: The origin of species. John Murray, London.
Darwin, C.R. 1862: On the various contrivances by which British and foreign orchids are fertilised by
insects. John Murray, London.
Darwin, C.R. 1881: The formation of vegetable mould through the action of worms, with
observations on their habits. John Murray, London.
Daugherty, C.H.; Towns, D.R.; Atkinson, I.A.E.; Gibbs, G.W. 1990: The significance of the biological
resources of New Zealand islands for ecological restoration. Pp. 9–21 in Towns, D.R.;
Daugherty, C.H.; Atkinson, I.A.E. (Eds). Ecological restoration of New Zealand islands.
Conservation Sciences Publication No. 2. Department of Conservation, Wellington.
Davis, M.R.; Lang, M.H. 1991: Increased nutrient availability in topsoils under conifers in the South
Island high country. New Zealand Journal of Forestry Science 21: 165–179.
de Maynadier, P.; Hunter, M.L. 1994: Keystone support. BioScience 44: 2.
Devries, M.F.W. 1995: Large herbivores and the design of large-scale nature reserves in Western
Europe. Conservation Biology 9: 25–33.
Dirzo, R.; Miranda, A. 1990: Contemporary neotropical defaunation and forest structure, function
and diversity—a sequel to John Terborgh. Conservation Biology 4: 444–447.
Druce, A.P. 1957: Botanical survey of an experimental catchment, Taita, New Zealand. DSIR
Bulletin 124. 81 p.
Ecroyd, C.E. 1996: The ecology of Dactylanthus taylorii and threats to its survival. New Zealand
Journal of Ecology 20: 81–100.
Ehrlich, P.R.; Dobkin, D.S.; Wheye, D. 1992: Birds in jeopardy. Stanford University Press, Stanford.
Esler, A.E. 1978: Botany of the Manawatu. Botany Division, DSIR. Government Printer, Wellington.
Estes, J.A. 1995: Top-level carnivores and ecosystem effects: questions and approaches. Pp. 151–158
in Jones, C.G.; Lawton, J.H. (Eds): Linking species and ecosystems. Chapman and Hall, New
York.
Estes, J.A.; Tinker, M.T.; Williams, T.M.; Doak, D.F. 1998: Killer whale predation on sea otters linking
oceanic and nearshore ecosystems. Science 282 (5388): 473–476.
Fahey, B.; Jackson, R. 1997: Hydrological impacts of converting native forests and grasslands to pine
plantations, South Island, New Zealand. Agricultural and Forest Meteorology 84: 69–82.
24
Payton et al.—Keystone species: the concept and its relevance in New Zealand
Fauth, J.E. 1999: Identifying potential keystone species from field data: An example from temporary
ponds. Ecology Letters 2: 36–43.
Fenner, M. 1998: The phenology of growth and reproduction in plants. Perspectives in Plant
Ecology, Evolution & Systematics 1: 78–91.
Franklin, J.F. 1993: Preserving biodiversity: species, ecosystems or landscapes? Ecological
Applications 3: 202–205.
Franklin, J.F. 1994: Response. Ecological Applications 4: 208–209.
Fugler, S.R. 1982: Infestations of three Australian Hakea species in South Africa and their control.
South African Forestry Journal 120: 63–68.
Gadgil, R.L. 1971: The nutritional role of Lupinus arboreus in coastal sand dune forestry 3. Nitrogen
distribution in the ecosystem before tree planting. Plant and Soil 35: 113–126.
Gautier-Hion, A.; Michaloud, G. 1989: Are figs always keystone resources for tropical frugivorous
vertebrates? A test in Gabon. Ecology 70: 1826–1833.
Gaze, P.D.; Clout, M.N. 1983: Honeydew and its importance to birds in beech forests of South Island,
New Zealand. New Zealand Journal of Ecology 6: 33–37.
Gitay, H.; Wilson, J.B.; Lee, W.G. 1996: Species redundancy—a redundant concept. Journal of
Ecology 84: 121–124.
Grant, P.J. 1989: A hydrologist’s contribution to the debate on wild animal management. New
Zealand Journal of Ecology 12: 165–169.
Greenwood, R.M. 1978: Rhizobia associated with indigenous legumes of New Zealand and Lord
Howe Island. Pp. 402–403 in Loutit, M.W.; Miles, J.A.R.(Eds): Microbial ecology. Springer
Verlag, Berlin.
Hanley, M.E.; Fenner, M.; Edwards, P.J. 1995: An experimental field study of the effects of mollusc
grazing on seedling recruitment and survival in grassland. Journal of Ecology 83: 621–627.
Heske, E.J.; Brown, J.H.; Guo, Q.F. 1993: Effects of kangaroo rat exclusion on vegetation structure
and plant species diversity in the Chihuahuan Desert. Oecologia 95: 520–524.
Holdaway, R.N. 1999: A spatio-temporal model for the invasion of the New Zealand archipelago by
the Pacific rat Rattus exulans. Journal of The Royal Society of New Zealand 29 (2): 91–105.
Holloway, J.T. 1959: Noxious-animal problems of the South Island alpine watersheds. New Zealand
Science Review 17: 21–28.
Howard-Williams, C.; Davies, J. 1988: The invasion of Lake Taupo by the submerged water weed
Lagarosiphon major and its impact on the native flora. New Zealand Journal of Ecology 11:
13–19.
Hughey, K.; Warren, A. 1997: Habitat restoration for wildlife nesting on degraded braided riverbeds
in New Zealand. Pp. 334–343 in Hale, P.; Lamb, D. (Eds): Conservation outside nature
reserves. Centre for Conservation Biology, University of Queensland, Brisbane, Australia.
Hurlbert, S.H. 1997: Functional importance vs keystoneness: reformulating some questions in
theoretical biocenology. Australian Journal of Ecology 22: 369–382.
Innes, J.; Hay, R.; Flux, I.; Bradfield, P.; Speed, H.; Jansen, P. 1999: Successful recovery of North
Island kokako Callaeas cinerea wilsoni populations, by adaptive management. Biological
Conservation 87: 201–214.
James, K.A.C.; Waghorn,G.C.; Powlesland, R.G.; Lloyd, B.D. 1991: Supplementary feeding of kakapo
on Little Barrier Island. Proceedings of the Nutrition Society of New Zealand 16: 93–102.
Jones, C.G.; Lawton, J.H.; Shachak, M. 1994: Organisms as ecosystem engineers. Oikos 69: 373–386.
Kaiser, J. 1998: Sea otter declines blamed on hungry killers. Science 282 (5388): 390–391.
Karnosky, D.F. 1979: Dutch elm disease: a review of the history, environmental implications,
control, and research needs. Environmental Conservation 6: 311–322.
Kelly, D.; Skipworth, J.P. 1984: Tradescantia fluminensis in a Manawatu (New Zealand) forest: 1.
Growth and effects on regeneration. New Zealand Journal of Botany 22: 393–397.
Science for Conservation 203
25
Kerley, G.I.H.; Whitford, W.G.; Kay, F.R. 1997: Mechanisms for the keystone status of kangaroo rats:
graminivory rather than granivory? Oecologia 111: 422–428.
Ladley, J.J.; Kelly, D. 1995: Explosive New Zealand mistletoe. Nature 378: 766.
Ladley, J.J.; Kelly, D. 1996: Dispersal, germination and survival of New Zealand mistletoes
(Loranthaceae): dependence on birds. New Zealand Journal of Ecology 20: 69–79.
Ladley, J.J.; Kelly, D.; Robertson, A.W. 1997: Explosive flowering, nectar production, breeding
systems, and pollinators of New Zealand mistletoes (Loranthaceae). New Zealand Journal
of Botany 35: 345–360.
Lambert, F.R.; Marshall, A.G. 1991: Keystone characteristics of bird-dispersed Ficus in a Malaysian
lowland rain forest. Journal of Ecology 79: 793–809.
Landres, P.B.; Verner, J.; Thomas, J.W. 1988: Ecological uses of vertebrate indicator species: a
critique. Conservation Biology 2: 316–328.
Lee, W.G.; Partridge, T.R. 1983: Rates of spread of Spartina anglica and sediment accretion in the New
River Estuary, Invercargill, New Zealand. New Zealand Journal of Botany 21: 231–236.
Lee, W.G.; Clout, M.N.; Robertson, H.A.; Wilson, J.B. 1990: Avian dispersers and fleshy fruits in New
Zealand. Acta XX Congressus Internationalis Ornithologici 3: 1617–1623.
Lloyd, B.D.; Powlesland, R. 1994: The decline of kakapo Strigops habroptilus and attempts at
conservation by translocation. Biological Conservation 69: 75–85.
Lloyd, D.G. 1985: Progress in understanding the natural history of New Zealand plants. New Zealand
Journal of Botany 23: 707–722.
Lord, J.M. 1990: The maintenance of Poa cita grassland by grazing. New Zealand Journal of Ecology
13: 43–49.
Makepeace, W. 1985: Some establishment characteristics of mouse-ear and king devil hawkweeds.
New Zealand Journal of Botany 23: 91–100.
Maloney, R.F.; Keedwell, R.J.; Wells, N.J.; Rebergen, A.L.; Nilsson, R.J. 1999: Effect of willow
removal on habitat use by five birds of braided rivers, Mackenzie Basin, New Zealand. New
Zealand Journal of Ecology 23: 53–60.
Manning, J.C.; Goldblatt, P. 1997: The Moegistorhynchus longirostris (Diptera: Nemestrinidae)
pollination guild: Long-tubed flowers and a specialized long-proboscid fly pollination system
in southern Africa. Plant Systematics and Evolution 206: 51–69.
Markwell, T. J. 1999: Keystone species on New Zealand offshore islands. Unpublished Ph.D thesis.
Victoria University of Wellington. 124 p.
McDowall, R.M. 1968: Interactions of native and alien faunas of New Zealand and the problem of fish
interactions. Transactions of the American Fisheries Society 97: 1–11.
McDowall, R.M. 1987: Impacts of exotic fishes on the native fauna. Pp. 291–306 in Viner, A.B. (Ed.):
Inland waters of New Zealand. DSIR Bulletin 241.
McDowall, R.M. 1990: New Zealand freshwater fishes: a natural history guide. Heinemann Reed–
MAF publishing group, Auckland–Wellington.
McEwan, W.M. 1978: The food of the New Zealand pigeon (Hemiphaga novaeseelandiae
novoseelandiae). New Zealand Journal of Ecology 1: 99–108.
McSaveney, M.J.; Whitehouse, I.E. 1989: Anthropic erosion of mountainland in Canterbury. New
Zealand Journal of Ecology 12: 151–163.
Meffe, G.K.; Carroll, C.R. 1994: Principles of conservation biology. Sinauer, Sunderland,
Massachusetts.
Menge, B.A.; Berlow, E.L.; Blanchette, C.A.; Navarrvete, S.A.; Yamada, S.B. 1994: The keystone
species concept: Variation in interaction strength in a rocky intertidal habitat. Ecological
Monographs 64: 249–286.
Meurk, C.D.; Norton, D.A.; Lord, J.M. 1989: The effect of grazing and its removal from grassland
reserves in Canterbury. Pp. 72–75 in Norton, D.A. (Ed.): Management of New Zealand’s
natural estate. New Zealand Ecological Society Occasional Publication No. 1. 119 p.
26
Payton et al.—Keystone species: the concept and its relevance in New Zealand
Mills, J.A.; Lee, W.G.; Mark, A.F.; Lavers, R.B. 1980: Winter use by takahe (Notornis mantelli) of the
summer-green fern (Hypolepis millefolium) in relation to its annual cycle of carbohydrates
and minerals. New Zealand Journal of Ecology 3: 131–137.
Mills, L.S.; Soulé, M.E.; Doak, D.F. 1993: The keystone species concept in ecology and conservation.
BioScience 43: 219–224.
Milton, K. 1991: Leaf change and fruit production in six neotropical Moraceae species. Journal of
Ecology 79: 1–26.
Moller, H.; Tilley, J.A.V. 1989: Beech honeydew: Seasonal variation and use by wasps, honey bees,
and other insects. New Zealand Journal of Zoology 16: 289–302.
Moller, H.; Townsend, C.R.; Ragg, J.R.; Bannister, P. 1993: Invasions and environmental impacts of
new organisms: can they be predicted? University of Otago, Dunedin.
Morales, C.F.; Hill, M.G.; Walker, A.K. 1988: Life history of the sooty beech scale (Ultracoelostoma
assimile) (Maskell), (Hemiptera: Margarodidae) in New Zealand Nothofagus forests. New
Zealand Entomologist 11: 24–37.
Mosley, M.P. 1978: Erosion in the south-eastern Ruahine Range: its implications for downstream
river control. New Zealand Journal of Forestry 23: 21–48.
Naiman, R.J.; Johnston, C.A.; Kelley, J.C. 1988: Alteration of North American streams by beaver.
BioScience 38: 753–762.
Nason, J.D.; Herre, E.A.; Hamrick, J.L. 1998: The breeding structure of a tropical keystone plant
resource. Nature 391: 685–687.
Navarrete, S.A.; Menge, B.A. 1996: Keystone predation and interaction strength: Interactive effects
of predators on their main prey. Ecological Monographs 66: 409–429.
Norbury, D.C.; Norbury, G.L. 1996: Short-term effects of rabbit grazing on a short-tussock grassland
in central Otago. New Zealand Journal of Ecology 20: 285–288.
Norbury, G.L. 1999: Ecological consequences of rabbit haemorrhagic disease. Pp. 10–14 in Jarvis,
B.D.W. (Comp.): Proceedings of the Rabbit Control, RCD: Dilemmas and Implications
Conference, Wellington, New Zealand, 30–31 March 1998. The Royal Society of New
Zealand Miscellaneous Series 55. 120 p.
Paine, R.T. 1966: Food web complexity and species diversity. American Naturalist 100: 65–75.
Paine, R.T. 1969: A note on trophic complexity and community stability. American Naturalist 103:
91–93.
Paine, R.T. 1971: A short-term experimental investigation of resource partitioning in a New Zealand
rocky intertidal habitat. Ecology 52: 1096–1106.
Paine, R.T. 1974: Intertidal community structure. Experimental studies on the relationship between
a dominant competitor and its principal predator Oecologia 15: 93–120.
Paine, R.T. 1995: A conversation on refining the concept of keystone species. Conservation Biology
9: 962–964.
Paine, R.T.; Castillo, J.C.; Cancino, J. 1985: Perturbation and recovery patterns of starfish-dominated
intertidal assemblages in Chile, New Zealand, and Washington State. The American
Naturalist 125: 679–691.
Parkyn, S.M.; Rabeni, C.F.; Collier, K.J. 1997: Effects of crayfish (Paranephrops planifrons:
Parastacidae) on in-stream processes and benthic faunas: a density manipulation
experiment. New Zealand Journal of Marine and Freshwater Research 31: 685–692.
Partridge, T.R. 1987: Spartina in New Zealand. New Zealand Journal of Botany 25: 567–575.
Payton, I.J.; Allen, R.B.; Knowlton, J.E. 1984: A post-fire succession in the northern Urewera forests,
North Island, New Zealand. New Zealand Journal of Botany 22: 207–222.
Peres, C.A. 1994: Primate responses to phenological changes in an Amazonian terre firme forest.
Biotropica 26: 98–112.
Power, M.E.; Mills, L.S. 1995: The Keystone cops meet in Hilo. Tree 10: 182–184.
Science for Conservation 203
27
Power, M.E.; Tilman, D.; Estes, J.A.; Menge, B.A.; Bond, W.J.; Mills, L.S.; Daily, G.; Castilla, J.C.; Lubchenco,
J.; Paine, R.T. 1996: Challenges in the quest for keystones. Bioscience 46: 609–620.
Powlesland, R.G.; Lloyd, B.D.; Best, H.A.; Merton, D.V. 1992: Breeding biology of the kakapo
Strigops habroptilus on Stewart Island, New Zealand. Ibis 134: 361–373.
Prins, H.T.T.; van der Jeugd, H.P. 1993: Herbivore population crashes in east Africa. Journal of
Ecology 81: 305–314.
Redford, K.H. 1984: The termitaria of Cornitermes cumulans (Isoptera, Termitidae) and their role in
determining a potential keystone species. Biotropica 16: 112–119.
Rose, A.B.; Pekelharing, C.J.; Platt, K.H. 1992: Magnitude of canopy dieback and implications for
conservation of southern rata–kamahi (Metrosideros umbellata–Weinmannia racemosa)
forests, central Westland, New Zealand. New Zealand Journal of Ecology 16: 23–32.
Rose, A.B.; Platt, K.H.; Frampton, C.M. 1995: Vegetation change over 25 years in a New Zealand
short-tussock grassland: effects of sheep grazing and exotic invasions. New Zealand Journal
of Ecology 19: 163–174.
Russell, G.H. 1994: Values and threats of willows in waterways – an engineer’s perspective. Pp. 23–
32 in West, C.J. (Comp.): Wild willows in New Zealand: proceedings of a Willow Control
Workshop hosted by Waikato Conservancy, Hamilton, 24–26 November 1993. Department
of Conservation, Wellington.
Sears, P.D. 1960: Grass/clover relationships in New Zealand. Proceedings of the VIII International
Grassland Congress: 130–133.
Shrader-Frechette, K.S.; McCoy, E.D. 1993: Method in ecology. Strategies for conservation.
Cambridge University Press, Cambridge.
Simberloff, D. 1998: Flagships, umbrellas, and keystones: Is single species management passé in the
landscape era? Biological Conservation 83: 247–257.
Smale. M.C. 1990: Ecological role of buddleia (Buddleja davidii) in streambeds in Te Urewera
National Park. New Zealand Journal of Ecology 14: 1–6.
Stapp, P. 1998: A re-evaluation of the role of prairie dogs in Great Plains grasslands. Conservation
Biology 12: 1253–1259.
Stewart, G.H.; Burrows, L.E. 1989: The impact of white-tailed deer Odocoileus virginianus on
regeneration in the coastal forests of Stewart Island, New Zealand. Biological Conservation
49: 275–293.
Strong, D.R. 1992: Are trophic cascades all wet? Differentiation and donor-control in speciose
ecosystems. Ecology 73: 747–754.
Tanner, J.E.; Hughes, T.P.; Connell, J.H. 1994: Species coexistence, keystone species, and
succession—a sensitivity analysis. Ecology 75: 2204–2219.
Terborgh, J. 1986a: Community aspects of frugivory in tropical forests. Pp. 371–384 in Estrada, A.;
Fleming, T.H. (Eds): Frugivores and seed dispersal. Junk, Dordrecht.
Terborgh, J. 1986b: Keystone plant resources in the tropical forest. Pp. 330–344 in Soulé, M.E. (Ed):
Conservation biology II. Sinaur Association, Sunderland, Massachusetts.
Thomas, C.D.; Moller, H.; Plunkett, G.M.; Harris, R.J. 1990: The prevalence of introduced Vespula
vulgaris wasps in a New Zealand beech forest community. New Zealand Journal of Ecology
13: 63–72.
Torrey, J.G. 1978: Nitrogen fixation by actinomycete-nodulated angiosperms. Bioscience 28: 586–591.
Towers, D. 1996: Diet overlap between co-existing populations of native blue ducks and introduced
trout: assessing the potential for competition. Unpublished PhD thesis. Massey University,
Palmerston North, New Zealand.
Townsend, C.R.; Crowl, T.A. 1991: Fragmented population structure in a native New Zealand fish:
an effect of introduced brown trout? Oikos 61: 347–354.
Tracy, C.R.; Brussard, P.F. 1994: Preserving biodiversity: species in landscapes. Ecological
Applications 42: 205–207.
28
Payton et al.—Keystone species: the concept and its relevance in New Zealand
Van Schaik, C.P.; Terborgh, J.W.; Wright, S.J. 1993: The phenology of tropical forests: Adaptive
significance and consequences for primary consumers. Annual Review of Ecology &
Systematics 24: 353–377.
Vitousek, P.M. 1990: Biological invasions and ecosystem processes: towards an integration of
population biology and ecosystem studies. Oikos 57: 7–13.
Wardle, J.A. 1984: The New Zealand beeches: ecology, utilisation and management. New Zealand
Forest Service, Christchurch. 447 p.
Webb, C.J.; Kelly, D. 1993: The reproductive biology of the New Zealand flora. Trends in Ecology
and Evolution 8: 442–447.
West, C.J. 1992: Ecological studies of Clematis vitalba (old man’s beard) in New Zealand. DSIR Land
Resources Vegetation Report No. 736.
Whitaker, A.H. 1987: The roles of lizards in New Zealand plant reproductive strategies. New Zealand
Journal of Botany 25: 315–328.
Williams, P.A.; Timmins, S.M. 1990: Weeds in New Zealand protected natural areas: a review for the
Department of Conservation. Science and Research Series 14, Department of Conservation,
Wellington, New Zealand. 114 p.
Williams, P.A.; Karl, B.J. 1996: Fleshy fruits of indigenous and adventive plants in the diet of birds in
forest remnants, Nelson, New Zealand. New Zealand Journal of Ecology 20: 127–145.
Willson, M.F.; Halupka, K.C. 1995: Anadromous fish as keystone species in vertebrate communities.
Conservation Biology 9: 489–497.
Wootton, J.T. 1994: Predicting direct and indirect effects: an integrated approach using
experiments and path analysis. Ecology 75: 151–165.
Wright, S.J.; Gompper, M.E.; DeLeon, B. 1994: Are large predators keystone species in neotopical
forests? The evidence from Barro Colorado Island. Oikos 71: 279–294.
Science for Conservation 203
29